Visualize the authorship DNA of open source software.
AI vs Human is a modern dashboard that analyzes GitHub repositories to reveal the real breakdown of contributions. We distinguish between human handcrafted code, AI-assisted commits (Copilot, Claude Code, Cursor, etc.), and automated maintenance bots.
- Granular AI Attribution: Detects specific tools like GitHub Copilot, Claude Code, Cursor, Aider, Devin, OpenAI Codex, and Gemini.
- Activity Timelines: A full year of activity visualized through a high-density contribution heatmap.
- Developer Ranks: Gamified badges from "Organic Architect" to "Digital Overseer" based on your contribution DNA.
- Lines of Code Analysis: Goes beyond commit counts to measure actual volume of code added by AI vs Humans.
- Shareable Identity: Custom generated scorecard images and embeddable README badges.
- Framework: Next.js (App Router)
- Database/Backend: Convex (Real-time sync & background processing)
- Styling: Tailwind CSS
- Visuals: Satori for OG images, Lucide & Simple Icons for icons.
- Validation: Biome for linting & Vitest for unit testing.
Our analysis engine uses a priority-based cascade to classify every commit:
- Author Type: GitHub API signals for Bot accounts.
- Author/Committer Data: Matching login, name, and email against 50+ known patterns.
- Co-Authorship: Parsing
Co-authored-bytrailers for AI tool signatures. - Message Markers: Scanning commit messages for "Generated with..." or tool-specific prefixes.
- PR Metadata: Inspecting the source PR of squash-merges to identify the original AI agent creator.
We welcome contributions! Whether it's adding new AI tool detection patterns or improving the UI, check out our Contributing Guide to get started.
Distributed under the MIT License. See LICENSE for more information.
Thanks goes to these wonderful people (emoji key):
 David Dias 💻 🎨 🤔 |
This project follows the all-contributors specification. Contributions of any kind welcome!